Google Dialogflow CX Multi-languages support in VoiceBot

[ilabs_support]

Hi Support,

I would like to ask the multi-languages supported in Google Dialogflow CX. I have a testing flow which included three languages. Each languages with have some TTS messages, intents, entity, etc. 

Since I am design to use voicebot to let user call in and choosing which language they want to listen. (like language 1, press 1, language 2, press 2.. or direct use some intent/entity to capture which language which caller is familiar with).

The flow for different languages is the same but the entity (training phrases) is different for different languages. How can I manage and control this? I would like Google Dialogflow CX can first capture the language which user chosen, then use the correct language for those speech recognition (STT) and Text-to-speech (TTS).

Thanks.

[Yash-Kavaiya]

For Voice bot development make sure you create your agent in global location 

You can use SSML ,DTML Functionality for voice bot 

For more best practice to create a voicebot Voice agent design best practices  |  Dialogflow CX  |  Google Cloud 

[ibaui]

Hi @ilabs_support,

Welcome to Google Cloud Community!

Dialogflow CX supports a wide range of languages. For a complete list, refer to the languages reference page. You can filter the language table to see which languages are supported for Speech-to-Text (STT) and Text-to-Speech (TTS).

Here's how you can manage and control multiple languages in your Dialogflow CX agent, including capturing the user's preferred language and using the correct language for STT and TTS:

Create Language-Specific Flows

Separate Flows: Create separate flows for each language you want to support. For example, you might have an "English Flow," a "Spanish Flow," and a "French Flow."

Language-Specific Content: Within each flow, include all the necessary components for that language:

• Intents: Define intents for each language to capture user input and trigger specific actions.
• Entities: Create entities for each language to recognize and extract relevant information from user utterances.

Utilize Parameterization

• Dynamic Parameters: Use parameters in your intents and responses to customize the behavior based on the user's language choice.
• Conditional Responses: Employ conditional responses to tailor the agent's output to the selected language.

Implement Language Detection and Switching:

• Intent Matching: Use intent matching to determine the user's preferred language based on initial utterances.
• Session Parameters: Store the detected language as a session parameter to influence subsequent interactions.
• Conditional Routing: Route the conversation to the appropriate language-specific flow based on the session parameter.

Configure Multilingual TTS and STT:

• TTS Voices: Set up TTS voices for each supported language in your Dialogflow CX agent's settings.
• STT Models: Ensure that your STT models are trained on the relevant languages to accurately recognize user input.

You can read through this documentation regarding Voice agent design best practices.

I hope the above information is helpful.